12. M. Müller, A. Paarmann, R. Ernstorfer, Nat. Commun. 5, 5292 (2014).13. E. Najafi, T. D. Scarborough, J. Tang, A. Zewail, Science 347,

164–167 (2015).14. A. Yurtsever, R. M. van der Veen, A. H. Zewail, Science 335,

59–64 (2012).15. E. Quinonez, J. Handali, B. Barwick, Rev. Sci. Instrum. 84,

103710 (2013).16. L. Piazza et al., Nat. Commun. 6, 6407 (2015).17. Materials and methods are available as supplementary

materials on Science Online.18. A. Gliserin, M. Walbran, F. Krausz, P. Baum, Nat. Commun. 6,

8723 (2015).19. C. Kealhofer et al., Science 352, 429–433 (2016).20. A. Lassise, P. H. A. Mutsaers, O. J. Luiten, Rev. Sci. Instrum. 83,

043705 (2012).

21. M. T. Hassan, H. Liu, J. S. Baskin, A. H. Zewail, Proc. Natl. Acad.Sci. U.S.A. 112, 12944–12949 (2015).

22. L. J. Wong, B. Freelon, T. Rohwer, N. Gedik, S. G. Johnson, NewJ. Phys. 17, 013051 (2015).

23. N. I. Zheludev, Science 328, 582–583 (2010).24. A. V. Kildishev, A. Boltasseva, V. M. Shalaev, Science 339,

1232009 (2013).25. R. A. Shelby, D. R. Smith, S. Schultz, Science 292, 77–79 (2001).26. X. Ni, Z. J. Wong, M. Mrejen, Y. Wang, X. Zhang, Science 349,

1310–1314 (2015).27. D. Ehberger et al., Phys. Rev. Lett. 114, 227601

(2015).28. P. Baum, A. H. Zewail, Proc. Natl. Acad. Sci. U.S.A. 104,

18409–18414 (2007).29. A. Feist et al., Nature 521, 200–203 (2015).

ACKNOWLEDGMENTS

This work was supported by ERC and Munich-Centre for AdvancedPhotonics. We thank F. Krausz for awesome support and discussionsand acknowledge I. Zhuravlev (Vocord SoftLab) for help with GPUprogramming.

SUPPLEMENTARY MATERIALS

www.sciencemag.org/content/353/6297/374/suppl/DC1Materials and MethodsFigs. S1 to S4References (30–36)Movies S1 to S5

13 April 2016; accepted 23 June 201610.1126/science.aaf8589

WATER RESOURCES

Connections between groundwaterflow and transpiration partitioningReed M. Maxwell1* and Laura E. Condon2

Understanding freshwater fluxes at continental scales will help us better predict hydrologicresponse and manage our terrestrial water resources. The partitioning ofevapotranspiration into bare soil evaporation and plant transpiration remains a keyuncertainty in the terrestrial water balance.We used integrated hydrologic simulations thatcouple vegetation and land-energy processes with surface and subsurface hydrology tostudy transpiration partitioning at the continental scale. Both latent heat flux andpartitioning are connected to water table depth, and including lateral groundwater flow inthe model increases transpiration partitioning from 47 ± 13 to 62 ± 12%.This suggests thatlateral groundwater flow, which is generally simplified or excluded in Earth system models,may provide a missing link for reconciling observations and global models of terrestrialwater fluxes.

Evapotranspiration (ET) is the largest terres-trial water flux, typically accounting formore water than runoff and for about 60%of precipitation (1). It contributes a substan-tial portion of the global land-energy budget

(2) as latent heat (LH) flux, which affects regionalclimate (3). Evapotranspiration commonly refers tothe combination of all evaporation (E) from baresoil, water bodies, plant canopy, and sublimationfrom snow, and transpiration (T) through plantstoma during photosynthesis. Here we focus onthe partitioning of ET into bare soil E and plant T.Because T depends on plant processes, whereasE depends on shallow soil moisture and energyavailability, these two factors respond differentlyto physical drivers and stress. Disentangling thesefluxes over large scales is a key step towardimproved understanding and prediction of water-shed dynamics, especially when considering fu-ture stresses.Connectivity between the surface and the sub-

surface provides a fundamental control on water-energy fluxes and partitioning (4). Connections

between the water table and evapotranspirationhave been shown inmodel simulations (5–9) andobservations of regional systems (10). Althoughtheory to estimate and simulate evapotranspirationhas evolved much over past decades (11), lateralgroundwater flow has yet to be incorporated incontinental-scale partitioning estimates (12). Quan-tifying the role of groundwater is important, be-cause if partitioning is tied to water table depthand lateral flow, accurate predictions of futurewater availability will require a more detailedunderstanding of the underlying processes con-trolling groundwater surface water interactionsthan are currently included inmost Earth systemsimulators. Current research on partitioning relieson either isotope approaches or land surfacemodels(13). These are fundamentally different methods,but bothmake critical assumptions about ground-water contributions to T and simplify groundwatersurface water interactions. Discrepancies in par-titioning estimates remain, and some have sug-gested that itmaybe systematically underestimatedby current Earth system models (13, 14).We used a continental-scale integrated hydrol-

ogy model simulation to study the role that la-teral groundwater flow plays in evapotranspirationpartitioning (15). We used the ParFlow model(16, 17), which couples groundwater and surfacewater flow with vegetation processes and snowdynamics (7, 18) to solve a complete water and

energy balance [figs. S1, S2, and S4 (15)]. The do-main covers 6.3 million km2 and encompassesmajor river basins in North America [includingthe Mississippi, Colorado, and Ohio basins; seeFig. 1 and fig. S3 (15)]. Simulations were run forone water year at hourly resolution driven by re-constructed meteorology. Transient simulationswere initialized using the results of a prior steady-state model over the same domain (19) and ad-ditional transient model initialization [tables S1and S2 (15)]. Themodelwas compared to 1.2milliontransient observations of stream flow, ground-water, and snow, and was shown to match ob-served behavior [figs. S5 to S20 (15)].This simulation generated roughly 1.3 trillion

outputs over the 1-year period, covering all keycomponents of the water energy budget. Twovariables that exemplify hydrologic stores and land-energy fluxes—water table depth and LH flux (theenergy counterpart to ET)—show many scales ofdetail within these output fields (Fig. 1). For ex-ample, although groundwater is generally shal-lower in the more humid eastern region of thedomain and deeper in arid western regions, la-terally convergent flow drives local variability,creating shallow water tables in river valleys thatcan supply surface water export. We see similarpatterns in LH flux, large-scale trends also drivenby climate gradients, yet nested hillsllope-scalefluctuations persist.The integrated model we used provides the

ability to explicitly evaluate interactions betweenvariables that are excluded from other global ap-proaches (i.e., water table depth and individualland-energy flux components such as T and E).For example, the ratios of plant transpiration tototal evapotranspiration (T/ET) are calculated di-rectly from model outputs and compared to com-piled stand-scale (14) andglobal (4,20,21) isotopicallybased partitioning estimates [Fig. 2, A and B (15)].We see broad agreement between model simu-lations and estimates across scale; the domain-averaged T/ET of 62 ± 12% agrees with recentglobal isotopically derived estimates (4, 14, 20, 21).To determine the role of lateral groundwater

flow in these partitioning estimates, we performeda second simulation that is identical to the basecase but allowed only vertical water movement,with no topographic influences or lateral flow.This approach is similar to current practice inland surface models (15). This “no lateral flow”simulation resulted in a domain-averaged T/ET

SCIENCE sciencemag.org 22 JULY 2016 • VOL 353 ISSUE 6297 377

1Integrated GroundWater Modeling Center, Department ofGeology and Geological Engineering, Colorado School ofMines, Golden, CO, USA. 2Department of Civil andEnvironmental Engineering, Syracuse University, Syracuse,NY, USA.*Corresponding author. Email: rmaxwell@mines.edu

RESEARCH | REPORTSon M

arch 30, 2020

http://science.sciencemag.org/

Dow

nloaded from

of 47 ± 13%, suggesting that lateral groundwaterflow plays a substantial role in the partitioning ofevapotranspiration. This shift was seen system-atically across all vegetation types, resulting in alowerpredictionof standandglobal scale estimates[fig. S20 (15)]. This indicates that topographicallydriven lateral flow provides an additional watersource for transpiration in groundwater conver-gence zones.To evaluate the relationships between parti-

tioning and groundwater directly, we plotted theannual average LH flux, accumulated T and E,

and T/E ratio against water table depth for ninetree and shrub land cover types, colored by soiltype:more than 3.2million points in total (Fig. 2,B to E). The distinct patterns shown in all fourplots indicate a connection between water tabledepth and land-energy fluxes. Plots of LH, T, andE each follow logistic curves with three distinctregions: (i) an energy- or temperature-controlledregion at shallow water table depths, where la-teral groundwater flow stabilizes soil moistureand removes any water limitation; (ii) a deep dis-connectedgroundwater regionwhere surface fluxes

are dependent on precipitation (i.e., water-limited);and (iii) a groundwater-controlled region discussedin priormodeling (8) and observational (10) studies,where land-energy fluxes are correlated with watertable depth.Similar to T and E, T/E ratios (Fig. 2F) also

show a clear groundwater dependence; however,they do not follow the same logistic curve be-havior as their composite fluxes. Here amaximumT/E partitioning [up to 35 in some locations,which also agrees with recent global isotopicallyderived estimates of T/E (4)] occurs at medium

378 22 JULY 2016 • VOL 353 ISSUE 6297 sciencemag.org SCIENCE

Fig. 1. Simulation results display great spatial complexity. Plots of transient (water year 1985) annually averaged water table (WT) depth (A) and LH flux (B)demonstrate the great detail we see in this simulation.Water table varies with climatic region; it is generally deeper west of the aridity divide and shallower in theeast. Likewise, LH flux is also greater in lesswater-limited regions but exhibits larger values in river valleys,where the water table depths are locally shallow underwhat might otherwise be water-limited conditions.

Fig. 2. Simulated annually averaged histograms of evapotranspirationpartitioning compare favorably to observations and are related to watertable depth.Simulated T/ET ratios (A and B) plotted by land cover type agreewith stand-scale estimates and observations (14) and are bracketed byglobal isotopic estimates, isotopic 1 (4) and isotopic 2 (20, 21), shown hereas box plots spanning the ± mean 1 SD. Scatterplots of water table versus

LH flux (C), bare soil evaporation (D), and T (E) suggest that ground-water’s role in moderating these fluxes (7–9, 27) extends across ourcontinental-scale domain. When plotted as a function of water tabledepth, the T/E ratio (F) peaks at the middle of the groundwater criticaldepth range, suggesting that Tmay be as much as 35 times greater than Ebecause of lateral groundwater flow.

RESEARCH | REPORTSon M

arch 30, 2020

http://science.sciencemag.org/

Dow

nloaded from

groundwater depths (0.5 to 5m). We showed thatthis change in T/E is due to different depths toavailable water for the separateE and T processes.E interfaces at the shallow soil surface, whereas Tdraws water deeper from within the root zone(15). A simple conceptual hillslope model (Fig. 3)shows that both T and E follow a similar logisticcurve with respect to water table depth, but thedepth range over which E is sensitive to watertable depth is shallower than that for T. This shiftcreates a peak in T/E at groundwater depthswhere E is water-limited but T is not. This sug-gests that plants’ access to deeper water allowsthem to draw water after the shallow soil is dryand can shift the balance of T partitioning in a

manner not currently included in Earth systemmodels.These results indicate that integrated simu-

lations of the terrestrial hydrologic cycle can pro-vide insights into process interactions that arecurrently lacking fromother approaches, and theyfurther motivate the need to advance model de-velopment as well as observations in this area.For example, although the 1-km resolution usedfor this simulation is fine-scale over such a largeextent, there are always processes below the gridscale that might be better represented with in-creased resolution. Additionally, these simulationsdo not include any anthropogenic influences, suchas groundwater pumping, irrigation, surfacewater

diversions, reservoirs, and urbanization. We en-vision a road map where these assumptions aresystematically relaxed and the model is contin-uously reevaluated against all available observa-tions [fig. S6 (15)].Although there has been much recent debate

regardingET partitioning, our results suggest thatlateral groundwater flowprocesses, currently over-looked inglobal approaches,mayplayan importantrole in characterizing evapotranspiration at largescales. Our integrated hydrologic simulation overmost of the major river basins in continentalNorth America suggests that groundwater notonly moderates E and T over the continent, butmay also increase the partitioning of T substan-tially. This underscores the importance of includ-ing lateral groundwater flow and storage in Earthsystem models (22), as well as developing carefulmultiscale observations of land surface fluxes andwater table depth to better explore this relation-ship (12). Finally, the impact of groundwater onland-energy fluxes may have important impli-cations for atmospheric simulation (23, 24) andsuggests that these feedbacks may go both ways:Changes in water table depth due to widespreadworldwide pumping may have a profound influ-ence on land-energy fluxes and climate (25, 26).

REFERENCES AND NOTES

1. K. E. Trenberth, L. Smith, T. Qian, A. Dai, J. Fasullo,J. Hydrometeorol. 8, 758–769 (2007).

2. K. E. Trenberth, J. T. Fasullo, J. Kiehl, Bull. Am. Meteorol. Soc.90, 311–323 (2009).

3. S. I. Seneviratne, D. Lüthi, M. Litschi, C. Schär, Nature 443,205–209 (2006).

4. S. P. Good, D. Noone, G. Bowen, Science 349, 175–177 (2015).5. J. S. Famiglietti, E. F. Wood, Water Resour. Res. 30,

3061–3078 (1994).6. G. D. Salvucci, D. Entekhabi, Water Resour. Res. 31, 1725–1739

(1995).7. S. J. Kollet, R. M. Maxwell,Water Resour. Res. 44, W02402 (2008).8. R. M. Maxwell, S. J. Kollet, Nat. Geosci. 1, 665–669 (2008).9. L. E. Condon, R. M. Maxwell, S. Gangopadhyay, Adv. Water

Resour. 60, 188–203 (2013).10. J. Szilagyi, V. A. Zlotnik, J. Jozsa, Ground Water 51, 945–951

(2013).11. K. Wang, R. E. Dickinson, Rev. Geophys. 50, RG2005 (2012).12. P. D. Brooks et al., Water Resour. Res. 51, 6973–6987 (2015).13. S. J. Sutanto et al., Hydrol. Earth Syst. Sci. 18, 2815–2827

(2014).14. W. H. Schlesinger, S. Jasechko, Agric. For. Meteorol. 189–190,

115–117 (2014).15. Materials and methods are available as supplementary

materials on Science Online.16. S. J. Kollet, R. M. Maxwell, Adv. Water Resour. 29, 945–958

(2006).17. R. M. Maxwell, Adv. Water Resour. 53, 109–117

(2013).18. R. M. Maxwell, N. L. Miller, J. Hydrometeorol. 6, 233–247

(2005).19. R. M. Maxwell, L. E. Condon, S. J. Kollet, Geosci. Model Dev. 8,

923–937 (2015).20. S. Jasechko et al., Nature 496, 347–350 (2013).21. A. M. J. Coenders-Gerrits et al., Nature 506, E1–E2

(2014).22. M. P. Clark et al., Water Resour. Res. 51, 5929–5956

(2015).23. R. M. Maxwell, F. K. Chow, S. J. Kollet, Adv. Water Resour. 30,

2447–2466 (2007).24. X. Jiang, G. Y. Niu, Z. L. Yang, J. Geophys. Res. 114, D06109

(2009).25. I. M. Ferguson, R. M. Maxwell, Environ. Res. Lett. 7, 044022

(2012).26. M. F. P. Bierkens, Water Resour. Res. 51, 4923–4947 (2015).27. J. F. Rihani, R. M. Maxwell, F. K. Chow, Water Resour. Res. 46,

W12523 (2010).

SCIENCE sciencemag.org 22 JULY 2016 • VOL 353 ISSUE 6297 379

Fig. 3. The relationship between groundwater depth and land-energy fluxes for an idealized hillslope.On the left, groundwater (GW) is shallow and neitherTor E is water-limited, becauseof lateral groundwaterconvergence, which stabilizes soil moisture. On the far right, groundwater is deep and disconnected fromthe land surface, resulting in lower Tand E that are limited by precipitation. In the two center regions, Tor Evaries with groundwater depth, but E disconnects at shallower depths than T. As a result of thesebehaviors, T/E partitioning is sensitive to water table depth across both center regions. The peak in T/Ereflects the differential in response to water table changes in E as compared to T. The roots as drawn hereare not to scale and do not reflect potential changes in density due to water table depth, nor does thisfigure reflect the presence of subsurface heterogeneity.

RESEARCH | REPORTSon M

arch 30, 2020

http://science.sciencemag.org/

Dow

nloaded from

ACKNOWLEDGMENTSThis work was supported by the U.S. Department of Energy Officeof Science, Offices of Advanced Scientific Computing Research andBiological and Environmental Sciences IDEAS project. Allsimulations were made possible through high-performancecomputing support from Yellowstone (ark:/85065/d7wd3xhc)provided by the National Center for Atmospheric Research'sComputational and Information Systems Laboratory, sponsored bythe National Science Foundation. The authors declare no financialconflicts. We thank M. M. Forrester for development andpresentation of the graphics in figs. S1 and S6, L. Bearup for reviewcomments, A. Bandler for assistance with surface water dataprocessing, and D. Osei-Kuffuor for assistance with the analytical

Jacobian. We also thank the anonymous reviewers for theirconstructive comments. All simulation data, models, and inputs arearchived and are available. Specific versions of ParFlow arearchived with complete documentation and may be downloaded at(http://inside.mines.edu/~rmaxwell/maxwell_software.shtml) orchecked out from a commercially hosted, free SVN repository andGITHUB; r730 was the version used in this study. The input dataand simulations presented here are available as supplementarymaterials on Science Online. R.M.M. and L.E.C. designed the studyand prepared the input data; R.M.M. ran the numericalexperiments; L.E.C. developed the model-observation comparisonframework and metrics; and R.M.M. and L.E.C. synthesized theresults, made the figures, and wrote the manuscript.

SUPPLEMENTARY MATERIALS

www.sciencemag.org/content/353/6297/377/suppl/DC1Materials and MethodsSupplementary TextFigs. S1 to S20Tables S1 and S2References (28–68)Input DataModel Output

30 March 2016; accepted 22 June 201610.1126/science.aaf7891

MICROBIOME

Cospeciation of gut microbiotawith hominidsAndrew H. Moeller,1,2 Alejandro Caro-Quintero,3 Deus Mjungu,4

Alexander V. Georgiev,5,6 Elizabeth V. Lonsdorf,7 Martin N. Muller,8 Anne E. Pusey,9

Martine Peeters,10 Beatrice H. Hahn,11 Howard Ochman1*

The evolutionary origins of the bacterial lineages that populate the human gut are unknown. Herewe show that multiple lineages of the predominant bacterial taxa in the gut arose via cospeciationwith humans, chimpanzees, bonobos, and gorillas over the past 15million years. Analyses of strain-level bacterial diversitywithin hominid gutmicrobiomes revealed that cladesofBacteroidaceaeandBifidobacteriaceae have been maintained exclusively within host lineages across hundreds ofthousands of host generations. Divergence times of these cospeciating gut bacteria are congruentwith thoseofhominids, indicating thatnuclear,mitochondrial, andgutbacterial genomesdiversifiedin concert during hominid evolution.This study identifies human gut bacteria descended fromancient symbionts that speciated simultaneously with humans and the African apes.

Cospeciation is a hallmark of intimate andancient symbiotic relationships (1,2).Humansand other mammals host communities ofbacterial symbionts, which are essential fornormal postnatal development and adult

health. Gut bacterial community membershipand abundance profiles are shaped by host ge-netics (3) and evolutionary history (4, 5), but alsoby diet (6), geography (7), and medical interven-tion (8). External factors exert a strong influenceon the composition of gutmicrobial communities,which are assembled anew in each host genera-tion. It is unknown whether lineages of gut bac-

teria persist within individual host lineages overtimescales long enough to lead to cospeciation.Here, we tested whether gut bacteria residingwithin present-day humans are descended fromancestral bacterial symbionts that cospeciatedwithhumans and the African apes.To test for cospeciation between hominids and

their gut bacteria, we assessed the congruencebetweenhominid and bacterial phylogenetic trees.Although DNA sequence data sets are availablefor gut microbiomes of hominids (4, 5, 9, 10),these are based on short ribosomal RNA (rRNA)amplicons or shotgun-metagenomic data, whichlack the resolution required to detect codiversi-fication between bacterial and hominid lineages.Bacterial rRNA sequences diverge too slowly totrack diversification over the time scale of hom-inid evolution (11, 12), and shotgunmetagenomicsequencing doesnot reliably capture orthologousgenetic regions fromrelatedbacteria indifferent hostspecies. Therefore, we used an amplicon sequencingapproach that assays quickly evolving protein-coding regions in bacterial genomes (phyloTags)(13) to profile strain diversity within the gut mi-crobiomes of humans, chimpanzees, bonobos, andgorillas. This fine-scale resolution allows inferenceof the phylogenies of closely related bacterial lin-eages, thereby enabling tests for cospeciation be-tween gut bacteria and the Hominidae.We amplified a variable region of theDNAgyrase,

subunit B (gyrB) gene from bacteria present in

fecal samples collected from humans living inConnecticut, USA (Homo sapiens; n = 16); wildchimpanzees fromGombeNational Park, Tanzania(Pan troglodytes;n=47); wild bonobos from threefield sites in the Democratic Republic of the Congo(Panpaniscus;n= 24); andwild gorillas from twofield sites in Cameroon (Gorilla gorilla; n = 24)(table S1 and fig. S1).We usedmultiple sets of prim-ers, each designed to target one of three bacterialfamilies prominent in the gut microbiome (13):the Bifidobacteriaceae, the Bacteroidaceae, or theLachnospiraceae. Amplicons were sequenced onthe Illumina MiSeq platform, generating 4,578,632reads averaging41,249 reads per sample. Sequenceswere screened for quality inQIIME (14) and filteredto eliminate sequencing errors (15) (data files S1to S3), and the relative frequencies of Bacteroida-ceae, Bifidobacteriaceae, and Lachnospiraceaestrains recovered from each sample were recorded(tables S2 to S4). Phylogenetic analyses were per-formed separately for each bacterial family. Se-quenceswere alignedwithClustalWas implementedinMEGA6.0 (16), andmaximum-likelihood treeswere constructed with a general time-reversibleplus invariant sites (GTR+I) model of nucleotidesubstitution. Relative node ages of maximum-likelihood trees were estimated in BEAST (17).The phylogenetic relationships among the Bac-

teroidaceae strains from the Hominidae mirrorthe relationships of their host species (Fig. 1).Previous studies of the microbiota of the Homin-idae compared the overall composition of theirmi-crobial communities (i.e., beta diversity) (4, 5,9, 10),but our phylogenetic analysis allowed us to tracethe evolution of individual bacterial lineages. Werecovered three well-supported clades (i.e., cladesin which the relationships among the bacteriallineages derived from different host species weresupported by more than 50% of bootstrap rep-licates) that contain strains from more than twohost species (Fig. 1, A to C). The topology of eachof these clades indicates ancient cospeciationbetween the Bacteroidaceae and the Hominidae:Within each of these clades, strains recoveredfrom bonobos and chimpanzees form separatesister groups, which together form a clade that issister to either gorilla-derived (Fig. 1, A and B) orhuman-derived (Fig. 1C) strains.The dominant pattern of diversification has been

the parallel cospeciation ofmultiple Bacteroidaceaelineages with their hominid host lineages (Fig.1A). However, in one case along the branch lead-ing to Pan, two Bacteroidaceae lineages arose

380 22 JULY 2016 • VOL 353 ISSUE 6297 sciencemag.org SCIENCE

1Department of Integrative Biology, 2506 Speedway A5000,University of Texas at Austin, Austin, TX 78712, USA. 2MillerInstitute for Basic Research in Science, 2536 Channing Way,University of California, Berkeley, CA 94720, USA. 3Corpoicá C.ITibaitata, Santa Fe de Bogotá, Bogotá, Colombia. 4GombeStream Research Center, Kigoma, Tanzania. 5Department ofAnthropology, Northwestern University, 1810 Hinman Avenue,Evanston, IL 60208, USA. 6Department of Human EvolutionaryBiology, Harvard University, 11 Divinity Avenue, Cambridge, MA02138, USA. 7Department of Psychology and BiologicalFoundations of Behavior Program, LSP 261B, Franklin andMarshall College, Lancaster, PA 17603, USA. 8Department ofAnthropology, University of New Mexico, MSC01-1040,Albuquerque, NM 87131, USA. 9Department of EvolutionaryAnthropology, 101 Biological Sciences Building, Duke University,Durham, NC 27708, USA. 10Institut de Recherche pour leDéveloppement, University of Montpellier, 34394 Montpellier,France. 11Departments of Medicine and Microbiology, PerelmanSchool of Medicine, 409 Johnson Pavilion, University ofPennsylvania, Philadelphia, PA 19104, USA.*Corresponding author. Email: howard.ochman@austin.utexas.edu

RESEARCH | REPORTSon M

arch 30, 2020

http://science.sciencemag.org/

Dow

nloaded from

Connections between groundwater flow and transpiration partitioningReed M. Maxwell and Laura E. Condon

DOI: 10.1126/science.aaf7891 (6297), 377-380.353Science 

, this issue p. 377Sciencemodels.and lateral flow strongly affect transpiration partitioning, thus explaining the inconsistencies between observations andflow into an integrated hydrologic model simulation for the entire United States. The model showed that water table depth outputs are often inconsistent with stable isotope observations. Maxwell and Condon incorporated dynamic groundwaterLand surface models are used to understand the partitioning of these fluxes on a continental scale; however, model

Soil evaporation and plant transpiration together contribute a substantial proportion of terrestrial freshwater fluxes.Groundwater flow drives partitioning

ARTICLE TOOLS http://science.sciencemag.org/content/353/6297/377

MATERIALSSUPPLEMENTARY http://science.sciencemag.org/content/suppl/2016/07/27/353.6297.377.DC1

REFERENCES

http://science.sciencemag.org/content/353/6297/377#BIBLThis article cites 61 articles, 4 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.ScienceScience, 1200 New York Avenue NW, Washington, DC 20005. The title (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience

Copyright © 2016, American Association for the Advancement of Science

on March 30, 2020

http://science.sciencem

ag.org/D

ownloaded from